Fixing occupation numbers

Matthias Krack matthia... at psi.ch
Mon Aug 8 16:35:00 UTC 2016


Hi,

you can try the following settings based on your input

   &SCF
    ADDED_MOS 3
    &SMEAR
     #      LIST 1-3 1/3
     METHOD ENERGY_WINDOW
     WINDOW_SIZE 0.1
     FIXED_MAGNETIC_MOMENT -1
    &END SMEAR
    EPS_SCF 2.0E-5
   &END SCF

which should result in an almost degenerate 1/3 occupation for that 
specific case of an oxygen atom. The degeneracy will improve for a setup 
with a really isolated oxygen atom, e.g.by using a larger cubic cell (> 
(10A)^3). In your (cell) setup the oxygen atom is still interacting with 
its images and its environment is not isotropic.

Matthias


On Thursday, 4 August 2016 14:46:38 UTC+2, Michael Trumm wrote:
>
> Ahoi,
>
> i am trying to perform some atomistic calculations using fractional 
> occupation numbers. (in this example it would be the neutral oxygen atom)
> after i initialized them via &SMEAR the occupation looks kind of ok, but 
> during the run cp2k chooses one of the tripletts.
> hence my question. is there any way to fix the fractional occupation 
> during the whole run to get a fractional triplett with the
> beta p-orbitals occupied p_x 1/3  p_y 1/3 and p_z 1/3 ? 
>
> thanks for any help,
> Michael.
>
> OUTPUT:
>
>  ALPHA MO EIGENVALUES AND MO OCCUPATION NUMBERS AFTER SCF STEP 20
>
> # MO index          MO eigenvalue [a.u.]            MO occupation
>          1                     -0.917069                 1.000000
>          2                     -0.394950                 1.000000
>          3                     -0.394539                 1.000000
>          4                     -0.312729                 1.000000
>          5                      0.468611                 0.000000
>          6                      0.493158                 0.000000
> # Sum                                                     4.000000
>
>   Fermi energy:                -0.241121
>
>   HOMO-LUMO gap:                0.781340 =  21.26 eV
>
>
>
>  BETA MO EIGENVALUES AND MO OCCUPATION NUMBERS AFTER SCF STEP 20
>
> # MO index          MO eigenvalue [a.u.]            MO occupation
>          1                     -0.778066                 1.000000
>          2                     -0.268824                 1.000000
>          3                     -0.213094                 0.000000
>          4                     -0.211852                 0.000000
> # Sum                                                     2.000000
>
>
>
> INPUT:
>
> &GLOBAL
>   PROJECT test
>   RUN_TYPE ENERGY
> &END GLOBAL
>
> &FORCE_EVAL
>   METHOD QS
>   &DFT
>    LSD
>    MULTIPLICITY 3
>    &PRINT
>      &MO
>        OCCUPATION_NUMBERS
>      &END MO
>    &END PRINT
>    BASIS_SET_FILE_NAME BASIS_MOLOPT
>    POTENTIAL_FILE_NAME GTH_POTENTIALS
>    CHARGE  0
>    &SCF
>     ADDED_MOS 2
>     &SMEAR
>       LIST 1-3 1/3
>     &END SMEAR
>     EPS_SCF 2.0E-5
>    &END SCF
>    &XC
>     &XC_FUNCTIONAL PBE
>     &END XC_FUNCTIONAL
>    &END XC
>   &END DFT
>   &SUBSYS
>    &CELL
>      ABC 5.189 8.953 20.129
>      ALPHA_BETA_GAMMA 90 101.11 90
>      PERIODIC XYZ
>    &END CELL
>    &KIND O 
>      BASIS_SET DZVP-MOLOPT-GTH
>      POTENTIAL GTH-PBE-q6
>    &END KIND
>    &COORD
>   O        -0.9968591421        2.9869391388        4.8752526707
>    &END COORD
>   &END SUBSYS
> &END FORCE_EVAL
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20160808/55c3bd66/attachment.htm>


More information about the CP2K-user mailing list